libc++ has no std::numeric_limits<__Float16> implementation and treats f8_t as is_integral.

Bug for is_integral: https://github.com/llvm/llvm-project/issues/102767
Bug for numeric_limits: https://github.com/ROCm/composable_kernel/issues/1460
--- a/library/include/ck/library/utility/check_err.hpp
+++ b/library/include/ck/library/utility/check_err.hpp
@@ -146,7 +146,7 @@ check_err(const Range& out,
     bool res{true};
     int err_count  = 0;
     double err     = 0;
-    double max_err = std::numeric_limits<ranges::range_value_t<Range>>::min();
+    double max_err = NumericLimits<ranges::range_value_t<Range>>::Min();
     for(std::size_t i = 0; i < ref.size(); ++i)
     {
         const double o = type_convert<float>(*std::next(std::begin(out), i));
@@ -178,7 +178,8 @@ check_err(const Range& out,
 template <typename Range, typename RefRange>
 std::enable_if_t<(std::is_same_v<ranges::range_value_t<Range>, ranges::range_value_t<RefRange>> &&
                   std::is_integral_v<ranges::range_value_t<Range>> &&
-                  !std::is_same_v<ranges::range_value_t<Range>, bhalf_t>)
+                  !std::is_same_v<ranges::range_value_t<Range>, bhalf_t> &&
+                  !std::is_same_v<ranges::range_value_t<Range>, f8_t>)
 #ifdef CK_EXPERIMENTAL_BIT_INT_EXTENSION_INT4
                      || std::is_same_v<ranges::range_value_t<Range>, int4_t>
 #endif